JPH04224260A - Combustion condition detecting device for internal combustion engine - Google Patents

Abstract

PURPOSE:To increase detecting accuracy by controlling amplification factor so that as large signal as possible can be inputted within a range where overflowing to an A/D converter is not made. CONSTITUTION:Switch-over of amplification factor in an amplifying circuit 3 is made several times in a knock judgment zone in a knock detecting microcomputer 4 according to the peak value in one cycle of signals detected by a knock sensor 1. Thus optimum detected signal can be inputted into an A/D converter 4a within a limit where overflowing to it is not made, to detect the combustion condition in an internal combustion engine precisely.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a combustion state detection device for an internal combustion engine that detects the combustion state of an internal combustion engine.

0002

[Prior Art] 1 of a device for detecting the combustion state of an internal combustion engine
In the case of a knock detection device that detects knocks in an internal combustion engine, the knock detection process involves converting a signal from a knock sensor installed in the internal combustion engine into digital data using an A/D converter, and converting this digital data into a microprocessor. It is known to detect the occurrence of knock in an internal combustion engine using computer processing.

In this case, in order to accurately transmit the output signal from the knock sensor to the microcomputer, it is necessary to
It is necessary to input as large a signal as possible to the /D converter without overflowing. However, the level of the output signal from the knock sensor is greatly affected by the rotational speed of the internal combustion engine in which the knock sensor is installed. For example, when the rotational speed of the internal combustion engine is low, the output signal level from the knock sensor becomes extremely low. On the other hand, when the rotational speed of the internal combustion engine increases, the output signal level becomes extremely large. Therefore, the knock detection device is provided with an amplifier having a plurality of amplification factors for amplifying or attenuating this output signal.

Conventionally, switching control of the amplification factor in this amplifier has been performed based on the maximum value (knock intensity value) of the output signal from the knock sensor within the previous knock determination period (for example, 64-45967). In addition to detecting the combustion state of an internal combustion engine, an in-cylinder pressure sensor is used to detect the in-cylinder pressure of the internal combustion engine, and the in-cylinder pressure sensor signal is sent to an A/D converter via an amplifier circuit.
It is known to detect the occurrence of a misfire by converting the data and inputting it into a microcomputer (Japanese Unexamined Patent Publication No. 45750/1983).

[0005]

[Problem to be Solved by the Invention] However, in such a knock detection device, the output of the knock sensor signal varies greatly from ignition cycle to ignition cycle even under the same operating conditions, such as when the rotational speed of the internal combustion engine is constant. , the knock intensity value also changes irregularly accordingly. In particular, the knock sensor signal output when a knock occurs is approximately 16 times larger than when no knock occurs, and the knock sensor signal output varies irregularly and significantly even within the same knock determination section.

For this reason, even in the conventional amplification factor switching control device, it is necessary to set the amplification factor to a low value to prevent overflow, and effective knock detection has not been possible. Also, in devices that detect misfires from the cylinder pressure of an internal combustion engine, the signal of the cylinder pressure sensor is converted to an A/D converter.
When inputting to the converter, the amplification factor of the in-cylinder pressure sensor signal was set to be low to prevent overflow of the A/D converter, making it impossible to effectively detect a misfire in terms of accuracy.

The present invention has been made to solve the above problems, and provides an internal combustion engine equipped with an amplification factor switching device capable of inputting as large a signal as possible within a range without overflowing to an A/D converter. The purpose of the present invention is to provide a combustion state detection device for an engine.

[0008]

[Means for Solving the Problems] In order to achieve the above object, a combustion state detection device for an internal combustion engine according to the present invention, as shown in FIG. an amplification means having a plurality of amplification factors for amplifying or attenuating the combustion signal detected by the detection means; and an amplification means having a plurality of amplification factors that amplify or attenuate the combustion signal detected by the detection means; A technical means is adopted that includes a switchable amplification factor switching means and a combustion state determining means for determining the combustion state of the internal combustion engine based on the output of the amplifying means.

[0009]

According to the present invention, the combustion state of the internal combustion engine is detected, and the amplification factor of the amplification means is switched multiple times within one combustion section based on the detection result.

0010

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be explained in detail below based on embodiments shown in the drawings. FIG. 2 is an overall configuration diagram showing the configuration of the apparatus of this embodiment. In FIG. 2, reference numeral 1 denotes a knock sensor that is disposed in a cylinder block of an internal combustion engine (not shown) and serves as a combustion state detection means by detecting vibrations caused by knocking of the internal combustion engine. This is a filter that removes the components and extracts only the signal of the frequency component peculiar to the knock that occurs when a knock occurs.

Reference numeral 3 denotes an amplifying circuit serving as an amplifying means for amplifying the signal extracted by the filter 2. The amplifying circuit 3 is composed of a first amplifier 3a and a second amplifier 3b. A first knock signal is amplified only by the first knock signal, and a second knock signal is amplified by both the first and second amplifiers 3a and 3b. In this embodiment, the amplification factor of the circuit from which the second knock signal is obtained is set to be four times the amplification factor of the first amplifier 3a from which the first knock signal is obtained. The first and second knock signals obtained by the amplifier circuit 3 are transmitted to input ports YG1 and YG of an A/D converter of a knock detection microcomputer 4, which will be described later.
4 is input.

Reference numeral 4 denotes a knock detection microcomputer (hereinafter referred to as
(referred to as knock ECU), and the above input port YG
A/D converter 4a, A/D converting the knock signal inputted at YG1 and YG4 from an analog signal to a digital signal.
A CPU 4b performs arithmetic processing for knock determination based on the output signal of the converter 4a, a read-only storage unit ROM4c stores control programs and control constants necessary for the calculation, and a read-only storage unit ROM4c stores calculation data during operation of the CPU 4b. It has a temporary storage unit RAM4d for temporary storage.

Reference numeral 5 denotes various sensors for detecting the state of the internal combustion engine and the vehicle, such as a crank angle sensor, a pressure sensor, and a water temperature sensor. Reference numeral 6 indicates a knock determination signal from the knock ECU 4 and a detection signal from the various sensors 5. This is an internal combustion engine control microcomputer (hereinafter referred to as internal combustion engine ECU) that performs arithmetic processing to control fuel injection amount and ignition timing.

Reference numeral 7 denotes an ignition device that generates high voltage at optimal ignition timing based on a signal from the internal combustion engine ECU 6, and supplies the high voltage to a spark plug of the internal combustion engine (not shown); 8 also the internal combustion engine ECU 6. This is an injector that supplies fuel to the internal combustion engine at the optimal fuel injection amount based on signals from the internal combustion engine. In addition, in this embodiment, the amplification factor of the circuit from which the second knock signal is obtained in the amplifier circuit 3 is the first knock signal.
Amplifier 3
Although the amplification factor b is set, it does not particularly need to be 4 times.

Further, in this embodiment, the amplification circuit 3 which can switch the amplification factor into two stages is provided, but of course, it is also possible to use an amplifier circuit whose amplification factor can be switched in three or more stages. FIG. 3 shows an external interrupt routine, for example, the BTDC of an internal combustion engine.
This interrupt timing is set at 10°C (before top dead center), and this interrupt routine starts from step N00.

In step N10, the rotational speed of the internal combustion engine is calculated from the time from the previous interruption to the current interruption, and in step N20, an appropriate knock determination interval, for example BTDC10, is determined based on the rotational speed information calculated in step N10.
C.A to 90.degree. C.A is calculated, and in step N30, a timer is set so that a timer interrupt occurs at the start time of the knock determination section calculated in step N20.

In step N40, processing is executed to set the cylinder count to the cylinder to be subjected to knock determination, and the process proceeds to step N50. In step N50, the process returns to the main routine described later. FIG. 4 shows a timer interrupt routine, which starts from step T00 corresponding to the interrupt time set in the routine of FIG.
This routine is repeatedly executed within the knock determination interval set by the routine (for example, every time a quarter wavelength has passed after the knock signal waveform transitions from the - side to the + side).

In step T10, the port of the A/D converter 4a is set, and in step T20, the initial amplification factor of the amplifier circuit 3 is set. Note that the method for setting the initial amplification factor in step T20 will be explained in detail later. In step T30, the knock intensity value V is detected. The knock intensity value V is, for example, the maximum value Vadj (j=1, 2, . . . n) of the knock sensor signal as shown in FIG. 5, and the maximum value Vpeak therein.

In step T40, it is determined whether knocking has occurred. Here, for example, the local maximum value Vad of the knock sensor signal and the knock judgment value Vref are compared by a routine not shown, and the number C that satisfies Vad≧Vref is calculated.
PLS is counted and when CPLS is equal to or greater than a predetermined value, it is determined that there is a knock. In step T50, the first rounding process of Vad detected in step T30 and the switching process of the amplification factor of the amplifier circuit 3 are executed. In addition,
The process of step T50 will be explained in detail later.

In step T60, it is determined whether or not the knock determination period has ended. If the knock determination period has not ended, the process returns to step T30, and if it is determined that the knock determination period has ended, the process advances to step T70. , In step T70, the knock determination result is input to the internal combustion engine ECU 6. In step T80, a second Vad smoothing process is performed based on the following equation.

[0021]

[Formula 1] Vmean=Vmeanj-1 + (Vmad -Vm
eanj-1 )/4 Note that Vmad is a value calculated by the first rounding process described later. In the next step T90, a knock state detection process is performed, and in step T100, the process returns to the main routine. Note that step T90 is a process for updating the knock determination level, but this process method is described in detail in Japanese Patent Laid-Open No. 64-45967, and since it has little relation to the gist of the present invention, it will not be described here. Explanation will be omitted.

FIG. 6 shows in detail the process of step T20 in FIG. 4, which is a routine for setting the initial amplification factor of the knock sensor signal for the current knock detection period, and starts from step T200. In step T210, R
The final second rounded value Vmean in the previous knock determination section of the cylinder to be controlled this time is read from AM4d, and in step T220, this Vmean is combined with a predetermined value VG4 that is a reference level for switching the amplification factor. If it is determined that Vmean is greater than or equal to the predetermined value VG4, the process proceeds to step T230, and if it is determined that Vmean is not greater than or equal to the predetermined value VG4, the process proceeds to step T.
Proceed to 250.

In step T230, the input port of the A/D converter 4a is set to the YG1 side, and in step T240
Then, the cylinder-specific A/D flag XG1 indicating that it is set to the YG1 side of the input port is set to 1, and the process advances to step T270 to end this routine. On the other hand, in step T250, the input port of the A/D converter 4a is set to the YG4 side, and in step T260, the A/D flag for each cylinder
is set to 0, and the process proceeds to step T270, where this routine ends.

FIG. 7 shows the process of step T50 in FIG. 4, which constitutes the main part of the process related to the present invention, and starts from step T500. In step T510, the XG1 flag set by the process shown in FIG. 6 is detected, and it is determined whether the amplification factor is 1x (whether the A/D port is set to YG1),
If the XG1 flag is 1, it is determined that the amplification factor is 1 times, and the process proceeds to step T520, and if the XG1 flag is 0, it is determined that the amplification factor is 4 times, and the process proceeds to step T56.
Go to 0.

In step T520, the first rounding process is performed using the following equation.

[0026]

[Formula 2] Vmadj=Vmadj-1+(Vad×4-Vmad
j)/16 Note that in step T520, since the amplification factor for amplifying the knock sensor signal is 1, the knock intensity value Vad is multiplied by 4 in the above equation before the averaging process is executed.

In the next step T530, the knock intensity value V
ad and a predetermined value VG1, the knock strength value Vad
is less than the predetermined value VG1, the process proceeds to step T540, in which the input port of the A/D converter 4a is set to the YG4 side, and the process proceeds to step T550, in which the input port of the A/D converter 4a is set to the YG4 side.
1 flag is set to 0, the process proceeds to step T600, and the process ends. On the other hand, if the knock strength value Vad is not less than the predetermined value VG1, the process directly proceeds to step T600 and ends.

Further, if it is determined in step T510 that the XG1 flag is not 1, the process advances to step T560, and in step T560, a second averaging process is also performed using the following equation.

[0029]

[Math. 3] Vmadj=Vmadj-1+(Vad-Vmadj)
/16 In the next step T570, the knock strength value Vad is compared with a predetermined value VG4 which is a reference level for switching the amplification factor in the same way as VG1, and if the knock strength value Vad is equal to or higher than the predetermined value VG4, the process proceeds to step T580. , In step T580, the input port of the A/D converter 4a is set to the YG1 side, and the process proceeds to step T590, where the input port of the A/D converter 4a is set to the XG1 side.
The flag is set to 0 and the process proceeds to step T600 to end. On the other hand, if the knock strength value Vad is not equal to or greater than the predetermined value VG4, the process directly proceeds to step T600 and ends.

[0030] In this embodiment, the A/D port is YG4.
, that is, when the amplification factor of the amplifier circuit 3 is 4 times, the knock strength value Vad was multiplied by 4 in step T520 and then the smoothing process was executed.
The smoothing process may be performed after the knock intensity value Vad is multiplied by 1/4 in step T560 based on the time of 1. FIG. 8 is a flowchart showing the main routine of the operation executed in the knock ECU 4, in which step M
Start from 00.

In step M10, the R of the knock ECU 4 is
Initialization of AM4d, etc. is executed, and in step M20, a knock determination level Vref is calculated using the following equation.

[0032]

[Formula 4] Vref = (K+KC)×(Vmean+Vmos
) Here, K is a constant set based on the operating conditions of the internal combustion engine, and KC is a variable for correcting the knock determination level Vref, which is determined in step M40, which will be described later. Further, Vmean is the second rounded value of the knock intensity value V determined in step T80 in FIG. 4, Vmos is a constant for absorbing A/D conversion errors, and the value of Vmos is 0.
But that's fine.

In step M30, a failure of the knock sensor 1 is detected, and in step M40, the knock judgment level V is detected.
Set a variable KC to correct ref. In the next step M50, the main routine cylinder counter is updated, and the process returns to step M20. In addition, step M
The processing method in No. 40 is described in detail in Japanese Unexamined Patent Publication No. 45967/1983, and since it has little relevance to the present invention, the explanation thereof will be omitted here.

FIG. 9 illustrates the knock sensor signal characteristics processed by the apparatus of this embodiment. In addition, 1
The signals amplified in two ways, double and quadruple, are always YG1 and Y
Each is input to the G4 port. The solid line in FIG. 9 is a characteristic diagram for explaining the operation of this embodiment.
The characteristics when the CU switches and processes these two knock sensor signals with different amplification factors are shown. The broken line shows the characteristics of the knock sensor signal when amplified at a constant amplification rate of 4 times.

In FIG. 9, the initial amplification factor is set to 4 times, and as shown in FIG. 7, the knock strength values Vad1 to
Since Vad3 is smaller than the predetermined value VG4, the amplification factor is maintained at 4 times. Since the next knock intensity value Vad4 becomes larger than the predetermined value VG4, the knock intensity value Vad
4, the amplification factor is switched to 1.times., and the amplification factor is maintained at 1.times. until the knock strength value Vad8 becomes smaller than the predetermined value VG1. Next, at the knock strength value Vad9, since the previous knock strength value Vad8 is less than the predetermined value VG1,
The amplification factor can be switched to 4x.

Therefore, by setting the amplification factor based on the immediately preceding knock intensity value Vad, it is possible to switch the amplification factor multiple times within the same knock determination section, thereby preventing changes in combustion conditions, misfires, and knock occurrences. Overflow can be reliably prevented regardless of irregular fluctuations in the knock sensor signal due to the presence or absence of the knock sensor. Also, the predetermined value VG
By appropriately setting 1 and VG4, the knock sensor signal can be effectively amplified, and the dynamic range of each cylinder can be utilized to the fullest, so knock detection accuracy can be improved.

Next, FIG. 10 shows another embodiment of the process of step T50 in FIG.
Start from 1. Step T510 and Step T5
20 and step T560 are the same as the processes with the same reference numerals in the routine shown in FIG. 7, so their explanation will be omitted. In step T510, the amplification factor of the amplifier circuit 3 is determined to be 1, and after the first averaging process, the process proceeds to step T521, and in step T521, the knock intensity value Vad is determined to be 4 times the value for the reason already stated. A deviation ΔV from Vmadj determined in step T520 is calculated. This deviation ΔV is a value indicating whether the knock sensor signal is currently increasing or decreasing.

In step T522, Vth is calculated by adding the value obtained by multiplying the knock intensity value Vad by 4 and the deviation ΔV obtained in step T521, and in step T525, this Vth
compared with the judgment value VG10 for switching the amplification factor,
If Vth is smaller than VG10, processing for switching the amplification factor to 4 times is performed in step T540 and step T5.
50, and proceeds to step T601 to end. On the other hand, if Vth is larger than VG10 in step T525, the process directly advances to step T601 and ends.

Note that step T540 and step T
550 is the same as the process with the same reference numeral in the routine shown in FIG. 7, so the explanation will be omitted. Also, step T51
0, it is determined that the amplification factor of the amplifier circuit 3 is 4 times, and step T
After the first averaging process is performed in the same manner as in step 520, the process proceeds to step T561, and in step T561, the knock intensity value Va is
Deviation ΔV between d and Vmadj found in step T560
Calculate.

In step T562, Vth is calculated by adding the knock intensity value Vad and the deviation ΔV obtained in step T561, and in step T565, this Vth is compared with a judgment value VG40 for switching the amplification factor, and Vth is determined to be VG.
If the value is 10 or more, a process for switching the amplification factor to 1x is executed in step T580 and step T590,
The process advances to step T601 and ends. On the other hand, step T5
If Vth is smaller than VG40 at step T65, the process directly advances to step T601 and ends.

Note that step T580 and step T
590 is the same as the process with the same reference numeral in the routine shown in FIG. 7, so the explanation will be omitted. By the method described above, it may be determined whether the knock sensor signal is currently increasing or decreasing, and the amplification factor of the amplifier circuit 3 may be set based on this. Next, another embodiment of the present invention will be described based on the figures shown below.

FIG. 11 shows an output signal waveform of a cylinder pressure sensor which is a combustion state detection means for detecting the cylinder pressure of an internal combustion engine, with the horizontal axis representing the crank angle and the vertical axis representing the cylinder pressure. Further, the broken line in the figure shows the characteristic of the cylinder pressure sensor signal when a misfire occurs in the internal combustion engine, and the solid line shows the characteristic when no misfire occurs, that is, when complete combustion occurs. At this time, the misfire determination process is performed by A/D converting the magnitude of each predetermined crank angle Δθ (for example, 5°C A) in the output waveform of the cylinder pressure sensor signal, inputting it to the microcomputer, and executing it. .

Therefore, as in the above embodiment, A/
To prevent overflow in the D converter, the cylinder pressure sensor signal is input to the A/D converter via an amplifier circuit. Next, FIG. 12 is a flowchart illustrating the operation of the apparatus of the present invention, which starts from step S100 in response to a signal from a misfire determination section setting routine (not shown).

[0044] In step S110, an initial amplification factor is set. Here, the initial amplification factor is set to the highest magnification, and in this embodiment, the amplification factor is set to 4 times, and at the same time, the XG1 flag is set to 0. In step S120, the current in-cylinder pressure value Vad is read from the detection signal from the in-cylinder pressure sensor, which is a combustion state detection means (not shown), and in step T510, the XG1 flag is detected and it is determined whether the amplification factor is 1. Then, if the XG1 flag is 1, step S130
If the XG1 flag is 0, the process proceeds to step S140.
Proceed to.

In step S130, the cylinder pressure value Vad read in step S120 is multiplied by four to perform an integration process for detecting a misfire, and in step S140, the cylinder pressure value Vad is directly integrated. In this integration process, the cylinder pressure value Vad is set to be negative during the compression stroke of the cylinder, and the cylinder pressure value Vad is set to be positive during the expansion stroke.
It is described in detail in Japanese Patent No. 50, and the explanation here will be omitted.

When the process in step S130 is completed, the process advances to step T530, and when the process in step S140 is completed, the process advances to step T570. Here, the same reference numerals are given to the same processes as those shown in FIG. 7, and since the processing method is the same as that already explained in detail, the explanation here will be omitted. In step S150, it is determined whether or not it is the timing to read the cylinder pressure in step S120. In this embodiment, the cylinder pressure is read at every predetermined crank angle (for example, 5 degrees Celsius), and when the predetermined crank angle is reached, step Proceed to S160.

In step S160, it is determined whether the misfire detection section has ended or not, and if it has not ended, step S1
The process returns to step 20 and repeats the process described above. On the other hand, if the misfire has ended, in step S170, the integration processing results in steps S130 and S140 are compared with the misfire determination value to determine whether a misfire has occurred, and if it is determined that a misfire has occurred, the misfire occurrence flag is set to 1. Then, the process advances to step S200 and returns to the main routine. On the other hand, step S
If it is determined in step 170 that no misfire has occurred, step S
The process proceeds to step S20, where the misfire occurrence flag is set to 0.
Proceed to 0 and return to the main routine.

[0048] The misfire determination method is also described in detail in, for example, Japanese Patent Application Laid-Open No. 45750/1983, and its explanation will be omitted here. Here, by detecting the misfire occurrence flag set in step S180 and step S190, misfire occurrence diagnosis processing such as cutting off fuel to the cylinder where the misfire has occurred can be executed.

[0049] Also, step T530 and step T
In step 570, it is determined whether or not to switch the amplification factor by comparing the in-cylinder pressure value Vad with a predetermined value, but similarly to the process shown in FIG. 10, it is determined whether the detection signal from the in-cylinder pressure sensor is increasing. The amplification factor may be switched after determining whether there is a decreasing trend. FIG. 13 is a diagram showing the characteristics of the in-cylinder pressure sensor signal used to explain the processing of the routine in FIG. 12, and similar to that explained in FIG. The rate is set to 4 times, and the predetermined value VG
The value of 4 is maintained until the output signal value Vad6 exceeds 4.

Next, from the output signal value Vad6, a predetermined value V
The amplification factor is increased to 1 until Vad8 becomes smaller than G1, and thereafter the amplification factor is converted to 4. As mentioned above, even in the misfire detection device using the in-cylinder pressure sensor signal, by converting the amplification factor within one combustion section,
It becomes possible to use the dynamic range of the A/D converter to the maximum, and the accuracy of misfire detection can be improved.

[0051]

As described above, in the present invention, the amplification factor is switched multiple times within one combustion section based on the signal from the combustion state detection means that detects the combustion state of the internal combustion engine. In addition to preventing A/
This has excellent effects in that the dynamic range of the D converter can be used to the maximum extent and the detection accuracy of the combustion state of the internal combustion engine can be improved.

[Brief explanation of the drawing]

FIG. 1 is a diagram corresponding to claims of the present invention.

FIG. 2 is an overall configuration diagram showing the configuration of an embodiment of the present invention.

FIG. 3 is a flowchart illustrating the operation of the apparatus shown in FIG. 2;

FIG. 4 is a flowchart for explaining the operation of the apparatus shown in FIG. 2;

FIG. 5 is a waveform diagram of a knock sensor signal.

FIG. 6 is a flowchart for explaining the operation of the apparatus shown in FIG. 2;

FIG. 7 is a flowchart illustrating the operation of the apparatus shown in FIG. 2;

FIG. 8 is a flowchart for explaining the operation of the apparatus shown in FIG. 2;

9 is a characteristic diagram for explaining the operation of the device shown in FIG. 2. FIG.

FIG. 10 is a flowchart illustrating another operation of the apparatus shown in FIG. 2;

FIG. 11 is a characteristic diagram for explaining another embodiment of the present invention.

FIG. 12 is a flowchart for explaining the operation of another embodiment.

FIG. 13 is a characteristic diagram for explaining the operation of another embodiment of the present invention.

[Explanation of symbols]

1 knock sensor 2 filter 3 amplifier circuit 4 knock detection microcomputer (knock E
CU)

Claims (4)

[Claims] 1. Combustion state detection means for detecting a combustion state of an internal combustion engine; amplification means having a plurality of amplification factors for amplifying or attenuating a combustion signal detected by said combustion state detection means; and said combustion state detection means. amplification factor switching means capable of switching the amplification factor of the amplification means multiple times within one combustion section based on the detection result of the amplification means; and combustion state determination means for determining the combustion state of the internal combustion engine based on the output of the amplification means. A combustion state detection device for an internal combustion engine, comprising: 2. The combustion state detection means is a knock sensor that detects knock vibration of the internal combustion engine, and the amplification factor switching means is based on the magnitude of the local maximum value Vad during a predetermined period of the knock sensor signal waveform. 2. The combustion state detection device for an internal combustion engine according to claim 1, wherein the amplification factor is switched within one combustion section. 3. The combustion state detection means is an in-cylinder pressure sensor that detects the in-cylinder pressure of the internal combustion engine, and the amplification factor switching means is configured to: The combustion state detection device for an internal combustion engine according to claim 1, wherein the amplification factor is switched within one combustion section. 4. The amplification factor switching means determines whether the detection result of the combustion state detection means is an increasing trend or a decreasing trend, and changes the amplification factor within one combustion section according to this determination result. The combustion state detection device for an internal combustion engine according to claim 1, wherein the combustion state detection device for an internal combustion engine is configured to switch.